metadata

language:
  - en
license: mit
base_model: distil-whisper/distil-small.en
tags:
  - generated_from_trainer
datasets:
  - PolyAI/minds14
metrics:
  - wer
model-index:
  - name: Distil Whisper Small finetuned on PolyAI Minds14 English US.
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Speech Transcription in English from e-banking domain.
          type: PolyAI/minds14
          config: en-US
          split: train
          args: en-US
        metrics:
          - name: Wer
            type: wer
            value: 0.3318442884492661

Distil Whisper Small finetuned on PolyAI Minds14 English US.

This model is a fine-tuned version of distil-whisper/distil-small.en on the Speech Transcription in English from e-banking domain. dataset. It achieves the following results on the evaluation set:

Loss: 1.0182
Wer Ortho: 0.3371
Wer: 0.3318

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: constant_with_warmup
lr_scheduler_warmup_steps: 50
training_steps: 400
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.2325	3.57	100	0.6222	0.3557	0.3472
0.0196	7.14	200	0.8475	0.3757	0.3689
0.0014	10.71	300	0.9729	0.3630	0.3555
0.0006	14.29	400	1.0182	0.3371	0.3318

Framework versions

Transformers 4.36.0
Pytorch 2.0.0
Datasets 2.1.0
Tokenizers 0.15.0